Post by: lagerfeldt on May 29, 2006, 09:05:11 AM
Yesterday I did a SRC from 48 to 44.1 using the Core Audio driver, and it struck me how bad it sounds, although I can't exactly put my finger on it.. "smeared" would be the word.
I've decided to get another program just for sample rate conversion, so which is the best on Mac?
Bias Peak?
BTW Peak looks like a damn nice program, maybe that's me switching from WaveBurner.
Post by: Andy Krehm on May 29, 2006, 09:24:32 AM
| lagerfeldt wrote on Mon, 29 May 2006 09:05 |
I did a little digging and being way too lazy to go thru the huge pile of posts I did a new one... Yesterday I did a SRC from 48 to 44.1 using the Core Audio driver, and it struck me how bad it sounds, although I can't exactly put my finger on it.. "smeared" would be the word. I've decided to get another program just for sample rate conversion, so which is the best on Mac? Bias Peak? BTW Peak looks like a damn nice program, maybe that's me switching from WaveBurner. |
Contact Daniel Weiss to see if he has ported Saracon over to the Mac yet. He has been working on it for a few months and must be just about finished.
Judging from reviews by PC owners, I believe that will be your best bet. If you can't wait, I think the one made by Antares, at half the price, would be your next choice (sorry can't remember the name).
Andy,
Silverbirch Productions
Post by: lagerfeldt on May 29, 2006, 10:31:03 AM
I remember a test posted here also that showed Bias Peak Pro 5 as being the top dog, or one of them at least.
This test shows Barbabatch come out on top, however the Peak version here is only 4, not the new 5
http://www.audioease.com/Pages/BarbaBatch4/Barba4SRCTest.htm l
Anyone know a newer comparison with Barbabatch 4 and Peak Pro 5?
Post by: Barry Hufker on May 29, 2006, 10:34:58 AM
Barbabatch from Audio Ease is excellent as well.
I own both programs and do my conversions with Barbabatch. They always sound good.
Barry
Post by: Bob Boyd on May 29, 2006, 10:46:00 AM
Barbabatch is solid and efficient when I need it.
Outside of conversion, they are very different programs intended for different purposes.
Based on the results I get on the SFC2, Saracon will be the king of the hill.
Post by: lagerfeldt on May 29, 2006, 10:59:41 AM
Post by: Jerry Tubb on May 29, 2006, 02:13:02 PM
| lagerfeldt wrote on Mon, 29 May 2006 08:05 |
Yesterday I did a SRC from 48 to 44.1 using the Core Audio driver, and it struck me how bad it sounds, although I can't exactly put my finger on it.. "smeared" would be the word. I've decided to get another program just for sample rate conversion, so which is the best on Mac? |
Hello Holger-
Yes the SRC in Core Audio via WaveBurner is less than desirable in many instances.
I've been using BarbaBatch for some time as well, sounds very good, although I can still hear a slight color change.
I tried Peak 5 SRC, but gave up after a few minutes of waiting, hit cancel.
(or you can use two DAWs, and capture at 44.1k after analog path.)
Best Regards
Post by: Daniel Weiss on May 29, 2006, 04:48:34 PM
| Andy Krehm wrote on Mon, 29 May 2006 15:24 |
Contact Daniel Weiss to see if he has ported Saracon over to the Mac yet. He has been working on it for a few months and must be just about finished. Judging from reviews by PC owners, I believe that will be your best bet. If you can't wait, I think the one made by Antares, at half the price, would be your next choice (sorry can't remember the name). Andy, Silverbirch Productions |
Not quite done yet with the mac version. But it is progressing well.
Daniel
Post by: Luke Fellingham on May 30, 2006, 05:55:27 AM
| Daniel Weiss wrote on Mon, 29 May 2006 21:48 | ||
Not quite done yet with the mac version. But it is progressing well. Daniel |
Do you have any predictions on time scale weeks, months or years?
Post by: lagerfeldt on May 30, 2006, 07:02:07 AM
| Daniel Weiss wrote on Mon, 29 May 2006 22:48 | ||
Not quite done yet with the mac version. But it is progressing well. Daniel |
Estimated price tag? And when are you doing a de-esser in AU/VST format, thanks
Post by: Bob Boyd on May 30, 2006, 09:59:52 AM
The "lite" version can be upgraded by paying the difference.
Reletively, it feels expensive for sample rate conversion but considering you're getting the same quality as the $3900 SFC2, it's worth it.
Post by: Jerry Tubb on May 31, 2006, 04:43:08 AM
| Daniel Weiss wrote on Mon, 29 May 2006 15:48 | ||
Not quite done yet with the mac version. But it is progressing well. |
That's good news Daniel, software developers need to remember us Mac guys.
There's a zillion of us in music & audio.
Look forward to doing a Saracon vs Barbabatch shootout. 48/96k to 44.1kHz.
BTW I just grokked the name Saracon...
Sa = sampling
Ra = rate
Con = conversion
Best Regards JT
Post by: bblackwood on May 31, 2006, 07:37:14 AM
| Bob Boyd wrote on Tue, 30 May 2006 08:59 |
Reletively, it feels expensive for sample rate conversion but considering you're getting the same quality as the $3900 SFC2, it's worth it. |
Yah, it's a heckuva deal. IIRC, Saracon actually outperforms (on 'paper') the SFC2 as well as everything else...
Post by: Matt_G on May 31, 2006, 11:04:42 AM
Looking at the Passband & Transition samples, tells a good story of the quality of the SRC's on test. Not conclusive evidence, as usual it's always best to use your ears, so I created some test files (unmastered & mastered snippets of music) & the 96kHz test files were SRC'd through a lot of the available converters. These include - Peak 4, Pro Tools Tweakhead, Audition, R8Brain Free, Sequoia, Wavelab, Barbabatch plus a couple more.
During the listening tests I conducted before the test files were made official. I couldn't believe my ears when I actually preferred the conversion done by Barbabatch. It sounded truer to the original 44.1kHz source file then the Saracon conversion did. The Saracon still sounded great, but it had added a slight increase in the high frequencies, which was confirmed by Alan Silverman late with his own blind tests. He upsampled a 44.1k test file to 96k & then SRC'd back down to 44.1 using Weiss Saracon & managed to pick Saracon's conversion over the original 44.1kHz test 8 out of 10 times due to the slight boost in HF content.
For the testing I did, I compared how the SRC'd files sounded against the original source file & listened for differences, not so much comparing the SRC'd files against each other. I just found that to my ears the Barbabatch converted file sounded truer to the original source then Saracon did. I also did some null tests & found a lot of the SRC'd test files nulled perfectly against each other, such as Adobe Audition, Pro Tools Tweakhead, Barbabatch 4 & R8Brain Free. All of these files nulled against each other. The ones that didn't null perfectly included Wavelab's SRC, Seqoiua & Saracon. There were some dreadful ones too like Peak 4, Waveburner & Logic which was an eye opener.
If you'd like to hear the listening tests that I compiled with the help of Dave Horrocks for the mastering webboard you can still download it here (to my surprise as it's about 6 months old now) http://www.iwserver2.com/LT.zip & the voting page is still up too http://src.infinitewave.ca/VOTE/vote.php
I am still hoping that iZotope will release their 64bit SRC as a standalone application for Mac as well. Up till then I think I'll stick with Barbabatch 4 even though BB4 is only 32bit & minimum phase. I have made requests to Aram from Audioease to make BB 64bit internal processing with some more dither options & they said it will be considered for a future release, but it's been over 6 months since I made that request & nothing more has happened to my knowledge.
Hope this helps you with your future choice.
Matt
Post by: Barry Hufker on May 31, 2006, 11:50:24 AM
Barry
Post by: Bob Boyd on May 31, 2006, 01:30:18 PM
| Jerry Tubb wrote on Wed, 31 May 2006 03:43 |
Look forward to doing a Saracon vs Barbabatch shootout. 48/96k to 44.1kHz. Best Regards JT |
Done this a million times. It won't even be close.
Post by: Matt_G on May 31, 2006, 06:34:59 PM
| Quote: |
Done this a million times. It won't even be close. |
Actually Barbabatch sounds more transparent to the original source file then Saracon, read my previous post.
Matt
Post by: Barry Hufker on May 31, 2006, 09:55:29 PM
Anecdotal evidence is fine for that individual, but only double blind listening tests and correctly performed measurements can determine which is "better."
Having said that, Barbabatch is better! So there!!
Barry
Post by: cerberus on May 31, 2006, 11:19:53 PM
uh...nine hundred dollars from your bank account?
at three times the price of what many are using now, (mine has cost me nothing since 2001), it ought be an order of magnitude better...do we need a double blind test to hear that?
i think dsp-quattro costs less than 200... btw; the programmer claims his src is as good sounding as peak 5 or barbatch.
jeff dinces
Post by: Bob Boyd on May 31, 2006, 11:29:14 PM
| Matt_G wrote on Wed, 31 May 2006 17:34 | ||
Actually Barbabatch sounds more transparent to the original source file then Saracon, read my previous post. Matt |
Interesting. It's been a while but I'll have to try it again. I know I tried it with Barbabatch 4 but I'm not sure what version it was.
Whenever I've upconverted 44.1 to 88.2 with Barbabatch, it always seemed to lose some top end detail comparitively.
I'm also using an SFC2, not Saracon so I'm not aware of what may have changed in the code.
Post by: Jerry Tubb on June 01, 2006, 04:37:19 AM
| Matt_G wrote on Wed, 31 May 2006 10:04 |
check out this website http://src.infinitewave.ca/.... It shows detailed passband & transition FFT's from the majority of SRC algorithms available today (both on the PC & the Mac) including Weiss Saracon, Barbabatch, Peak 5 etc. |
Very interesting graph comparisons... looks like they've updated them recently.
Saracon, Barbabatch, Peak 5, PT 7 Tweakhead, & Sonic HD look very good & pretty comparable. Of course it's the sound that counts... hearing is believing!
Noticed these are 96k to 44.1k SRC examples.
Is it possible that some softwares sound better on the upsampling, others on downsampling?
| Bob Boyd wrote on Tue, 30 May 2006 08:59 |
The "lite" version (stereo, 96k PCM) is $900. |
Looks like I'll be trying Saracon "Lite" : )
| Barry Hufker wrote on Wed, 31 May 2006 10:50 |
When this thread first appeared, I brought it to Audio Ease' attention. They said a 64 bit version is in development. |
Now that's interesting news, hopefully they'll have an upgrade path.
Looks like we're headed for a Saracon vs Barbabatch shootout at some point.
Cheers JT
Post by: Matt_G on June 01, 2006, 09:12:54 AM
| Barry Hufker wrote on Thu, 01 June 2006 11:55 |
While I am a Barbabatch fan -- and while I'll agree "what sounds good *is* good," to determine if product A is better than product B, there has got to be some solid, professional testing. Anecdotal evidence is fine for that individual, but only double blind listening tests and correctly performed measurements can determine which is "better." |
The way I see it, the goal of sample rate conversion should be to remain as transparent & invisible as possible. This has very little to do with comparing the sound of one SRC to another & everything to do with comparing each SRC to the original source file to check for transparency. Meaning it should sound as close to the source file as possible without any obvious colouration. Null testing can also help in the aid of this.
Matt
Post by: Arf! Mastering on June 02, 2006, 12:10:28 AM
Post by: dcollins on June 02, 2006, 12:56:48 AM
| AlanS wrote on Thu, 01 June 2006 21:10 |
No SRC is perfect and the differences between the best ones are very small. |
Unless used as an upsampler, then it's reported to be "better" than the source.
Potentially better than perfect!
Agreed about trying to null-test things like this.
DC
Post by: Barry Hufker on June 02, 2006, 03:33:16 PM
1. Using a Paqrat or Apogee PSX100 (or any similar device spreading two channels of 88.2k/24 bits across 8 tracks), record your music.
2. Because the stereo recording is laid across 8 tracks, the arrangement of the recording would be this: Even tho' the entire sampling frequency is 88.2k, the effective sampling rate for each track is 44.1k because of the bit splitting between tracks.
Track 1: Left Channel odd samples (Most Significant 16 bits,)
Track 2: Right Chanel odd samples (Most Significant 16 bits,)
Track 3: Left Channel even samples ""
Track 4: Right Channel even samples ""
Track 5: Left Channel odd samples ("bottome" 8 bits, equalling 24 bits when added to Track 1)
Track 6: Right Channel even samples ("bottom" 8 bits, equalling 24 bits when added to Track 2)
Track 7: Left Channel even samples ("bottom" 8 bits, equalling 24 bits when added to Track 3)
Track 8: Right channel odd samples ("bottom" 8 bits, equalling 24 bits when added to Track 1)
Normally after recording, one would play back thru the Paqrat or PSX100 to restore a stereo stream of 88.2k and 24 bits. Instead:
1. Pan tracks 1 and 3 left, with the faders at zero.
2. Pan tracks 2 and 4 right, with the faders at zero.
3. Pan tracks 5 and 7 left, but lower the fader 96 dB.
4. Pan tracks 6 and 8 right, but lower the fader 96 dB.
Listen to the music. You have effectively reduced the sampling frequency to 44.1k, while retaining all 24 bits. Because you haven't used an antialiasing filter, the sound should be more open than it would be otherwise because there are no phase shifts due to the antialiasing filter.
NOTE: Music with a great deal of high frequency content won't sound good with this method because of all the aliases. But classical and jazz should do well.
Barry
Post by: dcollins on June 02, 2006, 03:58:22 PM
| Barry Hufker wrote on Fri, 02 June 2006 12:33 |
Because you haven't used an antialiasing filter, the sound should be more open than it would be otherwise because there are no phase shifts due to the antialiasing filter. |
The filters never have any phase-shift as they are FIR's....
| Quote: |
NOTE: Music with a great deal of high frequency content won't sound good with this method because of all the aliases. But classical and jazz should do well. |
Why is this approach better? When the proper implementation works with any music? Or did he just like the sound of a little aliasing?
DC
Post by: Barry Hufker on June 02, 2006, 04:11:40 PM
I was working on an acoustic guitar project. When the guitarist came in, I had him listen to the two samples. All I said was is there one you like better than the other. I believe I played the Faulkner method first and then the conventional version. He immediately picked the first say it was more life-like and more natural sounding.
Give it a try.
Barry
Post by: jtr on June 04, 2006, 08:48:30 PM
I've always found Barbabatch to be a good general purpose SR utility- it does other things as well. Even if you eventually decide something else is more to your liking, Barbabatch is one of those utilities that (in Mac land) solves a ton of annoying file conversion problems. Sort of a Swiss Army Knife of conversion programs, and well worth the money.
You don't really begin to appreciate it until you get that one project for a multimedia developer or broadcaster where you need to convert 150 sound files from one format to another AND do SR conversion, or something else. The user interface is very good.
Post by: Bob Boyd on June 04, 2006, 09:15:04 PM
Post by: Jerry Tubb on March 22, 2007, 01:21:58 PM
http://www.audioease.com/Pages/BarbaBatch4/versionhistory.ht ml
I discovered it yesterday, was still using v4.0.29 from last year.
JT
Post by: Peter Beckmann on March 22, 2007, 01:47:43 PM
| Jerry Tubb wrote on Thu, 22 March 2007 17:21 |
Just thought I'd mention there's a new update for Barbabatch - v.4.0.36. http://www.audioease.com/Pages/BarbaBatch4/versionhistory.ht ml I discovered it yesterday, was still using v4.0.29 from last year. JT |
Thanks Jerry
Another Barbabatch user here since way way back. A really useful piece of software, sounds good and plus, the customer support has always been incredible.
How will the laid back Dutchmen fare head to head with the Swiss?
Bring on the battle of the SRCs!
PB
Post by: Matt_G on March 25, 2007, 10:51:41 AM
| Jerry Tubb wrote on Fri, 23 March 2007 03:21 |
Just thought I'd mention there's a new update for Barbabatch - v.4.0.36. http://www.audioease.com/Pages/BarbaBatch4/versionhistory.ht ml I discovered it yesterday, was still using v4.0.29 from last year. JT |
Yes I regularly visit to check on updates, I'm dissapointed that since this thread started there is still no sign of a 64bit SRC from Audioease.
It's also good to mention WaveEditor & SampleManager from audiofile engineering as a serious contender for no.1 SRC on the mac platform as it uses iZotope's 64bit SRC which is technically close to perfect. Audiofile Engineering also has plans to include all the options for the iZotope SRC for manual adjustments of the filters, slopes, pre-ringing etc.
The other good thing is that they are a lot cheaper then any of the competition & the customer service is 2nd to none, I get bug fixes & feature updates within 24hours most times
Matt
Post by: Ralf Kleemann on March 27, 2007, 12:21:42 PM
http://www.audiofile-engineering.com/sample_manager.php
I'm not affiliated with the developer in any way.
Best regards,
Ralf
Post by: Matt_G on March 27, 2007, 08:37:22 PM
| Ralf Kleemann wrote on Wed, 28 March 2007 02:21 |
Sample Manager 3 is currently on a 70$ offer. I got mine yesterday. Even if you just buy it for the batch processing, e. g. for mass-joining split stereo files, I think it's totally worth it. Together with Automator, you can build droplets that perform a specific task, which can save quite a bit of time. http://www.audiofile-engineering.com/sample_manager.php I'm not affiliated with the developer in any way. Best regards, Ralf |
I got my copy for $45, if you have a competing product such as Barbabatch or Peak they will give it to you for $45. In my opinion SampleManager is heaps better then Barbabatch in terms of SRC quality & the flexibility of the processing. It's an absolute must for mac users at this price. I'm not affiliated with them either, just sharing my experience with the other Mac users out there.
https://www.audiofile-engineering.com/shop/compUpgrades.php
https://www.audiofile-engineering.com/shop/eligibility.php
Matt
Post by: minister on March 27, 2007, 11:21:51 PM
| Matt_G wrote on Tue, 27 March 2007 19:37 |
I got my copy for $45, if you have a competing product such as Barbabatch or Peak they will give it to you for $45. In my opinion SampleManager is heaps better then Barbabatch in terms of SRC quality & the flexibility of the processing. |
Seriously? Heaps better? BB is pretty good.
did you do a side by side comparison on the same file?
can you post or e-mail identical short clips?
Post by: Barry Hufker on March 28, 2007, 12:35:19 AM
It is painfully slow but I start it and have it cook overnight so it's ready in the morning. Did I mention the quality is *really* good and I have been a huge Peak critic (and owner) for years.
Barry
Post by: Jerry Tubb on March 28, 2007, 01:20:18 AM
| Barry Hufker wrote on Tue, 27 March 2007 23:35 |
It is painfully slow but I start it and have it cook overnight so it's ready in the morning. |
Hey Barry,
I agree Peak 5 (level 5) SRC sounds really good, but the fact that is sooo sloooow totally disqualifies it for me. I usually don't have the luxury of leaving it to crunch overnight, usually the producer wants to take a ref disc that same day.
Its amazing to me that Peaks' SRC takes that long on a dual processor G5... what, 20 minutes per song? There's some serious number crunchin' power happening with a G5, leads me to wonder what's up with Peak, does anything else including Weiss take that long?
Anyway I'm still a fan of BarbaBatch, don't hear any grainy-ness happening there, although a slight but acceptable color change is evident.
Found Matt Gs' comments interesting regarding Sample Manager 3.
JT
Post by: Matt_G on March 28, 2007, 03:51:06 AM
| minister wrote on Wed, 28 March 2007 13:21 | ||
Seriously? Heaps better? BB is pretty good. did you do a side by side comparison on the same file? can you post or e-mail identical short clips? |
Hey Tom, you don't have to take my word for it just download the SampleManager demo & try it out for yourself. I've used Barbabatch for years & when I bought it there wasn't a whole lot of options available on the Mac in terms of batch conversion or quality SRC. Unfortunately for Audioease they haven't been keeping Barbabatch up to date with the competition lately. All the best software SRC's seem to be using 64bit precision these days such as Weiss Saracon & iZotope's SRC.
Matt
Post by: Andy Krehm on March 28, 2007, 11:58:50 AM
| Jerry Tubb wrote on Wed, 28 March 2007 01:20 | ||
Hey Barry, I agree Peak 5 (level 5) SRC sounds really good, but the fact that is sooo sloooow totally disqualifies it for me. I usually don't have the luxury of leaving it to crunch overnight, usually the producer wants to take a ref disc that same day. Its amazing to me that Peaks' SRC takes that long on a dual processor G5... what, 20 minutes per song? There's some serious number crunchin' power happening with a G5, leads me to wonder what's up with Peak, does anything else including Weiss take that long? Anyway I'm still a fan of BarbaBatch, don't hear any grainy-ness happening there, although a slight but acceptable color change is evident. Found Matt Gs' comments interesting regarding Sample Manager 3. JT |
Hey JT:
I'm testing Weiss Saracon and I just timed a 96k to 44.1 k conversion of a SD2 split file at 61.2 MB each side and it took 5.5 minutes.
I usually convert one file while working on another so this is not a problem and of course a 48 to 44.1 conversion averages out at less than 2 min.
I checked Saracon vs PTs vs. Wave Editor and be honest, I didn't find a really clear cut winner in every case. In fact, one song would seem to sound "better" with one SRC and then a different song sounded "better" with another SRC.
Unfortunately I don't have the time for really exhaustive testing, nor the interest really, b/c they were generally so close. My tests were a couple of months ago but I think I liked the Saracon more times than PTs but surprisingly, PTs was very close (and a lot faster)! Unfortunately my demo of Wave Editor expired before I could use it as much but I still have the same conclusion in that it wasn't remarkably better or worse.
Wave Editor also seemed to have more steps than PTs or Saracon which makes it a little more cumbersome to use but its also seems more flexible and I noticed that it has more functions that might come in handy.
Daniel Weiss said he is considering having a lower priced version without the DSD (I hope he doesn't mind me mentioned this) and if this is the case, I'll probably go with Saracon for my software SRC.
Post by: cerberus on March 28, 2007, 01:19:38 PM
jeff dinces
Post by: jdg on March 28, 2007, 01:31:11 PM
i've started transfering the files over to my mbp to src in peak, and then xfering them back to the G5 for sequencing.
i haven't done any offical benchmarks.
but about 20min per track on my G5
and about 35min for the whole album on my macbook pro.
Post by: Bob Boyd on March 29, 2007, 01:30:53 AM
| Andy Krehm wrote on Wed, 28 March 2007 10:58 |
I'm testing Weiss Saracon and I just timed a 96k to 44.1 k conversion of a SD2 split file at 61.2 MB each side and it took 5.5 minutes. |
I had this happen when I first tried Saracon too. You need to increase the buffers in the Preferences panel. The manual mentions this but may not stress it's importance.
The optimal settings will vary from computer to computer but try opening the first buffer all the way and the second one about 3/4. (I'm away from my studio computer so I don't remember what the faders are labeled.)
On my Dual 2.7 G5, a STEREO file going from 88.2k to 44.1/24 with pow-r will (at it slowest!) be over 2x realtime. Expect much faster results on the whole.
Post by: Bob Boyd on March 29, 2007, 01:33:28 AM
Post by: bigaudioblowhard on March 29, 2007, 02:23:12 AM
| Daniel Weiss wrote on Mon, 29 May 2006 14:48 |
... Not quite done yet with the mac version. But it is progressing well. Daniel |
At the bottom of your web page for Saracon, you claim to prefer the hardware unit.
Could you please breifly explain why?
bab
Post by: Matt_G on March 29, 2007, 02:55:52 AM
| cerberus wrote on Thu, 29 March 2007 03:19 |
i'm considering peak just for src; good things come to those who wait. jeff dinces |
Why??! I mean why waste $599 on a product that takes 45mins to SRC a 3 minute file when the quality is no better then the iZotope 64bit SRC in SampleManager which can process the same file in 5mins at a cost of $70?
I've used Peak in the past & I've never trusted it's sonic integrity as an audio editor, not to mention the endless bugs. Now that they release a decent SRC in version 5 everyone thinks it's great. I don't get it... Do yourself a favour & check out WaveEditor (2 track editor) or SampleManager (batch convertor). This is a company that has their head in the right place, excellent support, excellent products, excellent quality & excellent pricing.
Ok I'll get off my soap box now...
Matt
Post by: cerberus on March 29, 2007, 06:34:31 AM
didn't know they had a simpler app with the izotope src (which i still haven't heard.)
jeff dinces
Post by: Matt_G on March 30, 2007, 10:32:36 AM
| cerberus wrote on Thu, 29 March 2007 20:34 |
checked out wave editor: made hash out of 32 float files. didn't know they had a simpler app with the izotope src (which i still haven't heard.) jeff dinces |
Hey Jeff, when did you try WaveEditor? which version? It's up to 1.2.10 now. Initially they had a few bugs handling some file headers, but I think they've well & truly sorted that out as I haven't had any problems for a long while. SampleManager is up to 3.0.8 & it has had the iZotope MBIT dither & 64bit SRC since version 3.0. It's primarily a batch convertor.
Matt
Post by: cerberus on March 30, 2007, 03:14:43 PM
it was wave editor 1.1.x. i tried it in january. i guess it was the header issue, but i could not figure out how to fix it with soundhack or other apps that can re-write the header.
sample manager seems more suited for my needs, so i will be demoing it soon. thanks for the infos.
jeff
Post by: Roland Storch on March 30, 2007, 06:13:48 PM
http://www.sonicstudio.com/pdf/pr/iZotope_SSLLC_SRCMBit.pdf
Will do a comparison next week between soundBlade and Barbabatch.
Post by: Yannick Willox on March 31, 2007, 04:15:18 AM
> check out this website http://src.infinitewave.ca/....
and especially the r8brain PRO minimal phase against ALL of the others ?
???
Post by: Matt_G on March 31, 2007, 07:15:09 AM
| Yannick Willox wrote on Sat, 31 March 2007 18:15 |
Nobody checked the graphs: > check out this website http://src.infinitewave.ca/.... and especially the r8brain PRO minimal phase against ALL of the others ? ??? |
If you're looking at the graphs alone nothing beats the "iZotope 64bit steep, no aliasing"graphs. Sure the R8 Brain Pro one is a close 2nd but it's for PC only & we are discussing Mac only SRC options. The iZotope SRC is available on the Mac platform in the form of Sample Manager 3 & Wave Editor 1.2. At the moment there is only options to change the quality from low to high with a slider. Version 1.3 of Wave Editor (due soon) will feature all the advanced parameters for adjustment (see attached graph). They will also put the advanced feautures in Sample Manager 3.2 when it's released.
Matt
Post by: Ralf Kleemann on April 01, 2007, 06:27:43 AM
| Matt_G wrote on Sat, 31 March 2007 13:15 |
They will also put the advanced feautures in Sample Manager 3.2 when it's released. |
That's very nice to know, looking forward to the update. What I also like about the company is that they are open about forthcoming releases and features. At Apple, you would already have been sacked for posting a screenshot to a forum...
Now that you posted the screenshot, could you elaborate which of the sliders is doing what, and to what benefit?
Best regards,
Ralf
Post by: Barry Hufker on April 01, 2007, 03:57:11 PM
Barry
Post by: Bob Boyd on April 01, 2007, 09:26:14 PM
I'm watching iZotope and trying the Sample Manager demo but at this point in time, I'm using Weiss Saracon. Great SRC.
Post by: Matt_G on April 02, 2007, 10:34:40 AM
| Ralf Kleemann wrote on Sun, 01 April 2007 20:27 | ||
That's very nice to know, looking forward to the update. What I also like about the company is that they are open about forthcoming releases and features. At Apple, you would already have been sacked for posting a screenshot to a forum... Now that you posted the screenshot, could you elaborate which of the sliders is doing what, and to what benefit? Best regards, Ralf |
http://www.izotope.com/tech/src/ This covers the design info quite well.
Post by: Matt_G on April 02, 2007, 10:45:35 AM
| Bob Boyd wrote on Mon, 02 April 2007 11:26 |
The current SRC in sB is not iZotope's nor is it evidently the classic SRC that so many people refer to liking from the older system. While it has gotten some positive feedback, I find the current "built-in" SRC as utilitarian in nature. As mentioned before, Sonic has licensed iZotope's for implementation at some point. I'm watching iZotope and trying the Sample Manager demo but at this point in time, I'm using Weiss Saracon. Great SRC. |
It's funny Bob, I used to like soundBlade's current SRC (which isn't the iZotope at the moment). I still think that despite the aliasing it can sound great on certain material, can't explain the affect I hear but it kind of has a smooth soft sound which is great on some stuff.
The iZotope SRC on it's high setting in WaveEditor can be a touch on the bright side but is likely the steepness of the default filter setting. See the attached image of the 'high quality' default setting in the current WE 1.2 version. With these being adjustable in 1.3, I bet we will get an amazing sounding SRC suitable for any style of music. With iZotope products it's all about high quality but giving the consumer all the options to create just what you need. MBIT+ is another example of having many options.
Matt
Post by: Roland Storch on April 02, 2007, 06:33:20 PM
| Barry Hufker wrote on Sun, 01 April 2007 20:57 |
I have just done a sweep from 20 Hz. to 40kHz, at 96K and 24 bit. I then converted from 96k to 44.1 kHz at 24 bit. I used three methods: 1. Soundblade; Peak and Barbabatch. Of the three, Peak comes closest to the ideal of being "out" by 22050 Hz. It has a fairly smooth fade just before that frequency. Second is Barbabatch which extends a little more past 22050 with a lesser (not as steep) fade out. The worst plainly, simply, dramatically is the conversion by Soundblade. I don't believe this to be one by Izotope. I couldn't find that in my list of plug-ins if it did indeed come with the program. So the one written by Sonic is nothing less than terrible, extending well past 22050 and is sure to cause aliasing. Barry |
Barry, I told Oliver Masciarotte from Sonic Studio about your measurements and he allowed me to post his reply here in this thread:
"Our SRC is designed to sound good, not fall within a old textbook example of anti-aliasing. In brief, all anti-image (when creating a file) and anti-alias (when playing back) measures are designed to reduce the audible artifacts of sampling.
The ?old school? thinking was that the designer tried to provide a flat frequency response in the passband, sever cut in the stopband (usually with a high order, low pass filter) and to hell with the group delay. A more modern approach, which we use, is to provide both flat frequency and phase response in the passband, controlled roll?off in the stopband and minimize overall group delay as much as possible. So, response above the Nyquist frequency is not severely cut, it is controlled, with an appropriate roll?off that reduces artifacts while preserving both passband ripple and, equally as important, phase response.
An interesting sidenote is that the poster seems to not have done any listening tests, since he says, ? and is sure to cause aliasing??Well, had he actually listened to these three versions, he would find they all sound different and he should pick his method/tool based on how it sounds with music, not solely based on measurements with tones."
I don?t want to add something here as long as I didn?t compare Barbabatch and soundBlade?s SRC by hearing. Hopefully I can do the comparisons this week an will tell my impression here again.
Post by: cerberus on April 02, 2007, 09:08:22 PM
jeff dinces
Post by: Barry Hufker on April 02, 2007, 10:54:47 PM
Looking at phase, soundBlade doesn't do any better or worse than Barbatch or Peak. Looking at the graph of a 1kHz tone, as I read it, there's a "shitload"* of aliasing -- much more than the other two and much higher in level than the other two. And I do understand about frequency roll-off and group delay.
Again, I'm no expert (but I am a proud soundBlade owner)so if someone wants to educate me I would be sincerely appreciative. Otherwise, SB's roll-off is doing what I suggested, which I suggested before reading the graph.
Barry
* In my new measurement system, I have done away with both Imperial and Metric measurements. Everything, no matter its size, weight, mass, speed, etc. comes in "shitloads". The only exception is "shit," which comes in "buttloads".
Example: really good sample rate conversion may cost a shitload of money.
Post by: dcollins on April 02, 2007, 11:23:52 PM
| Barry Hufker wrote on Mon, 02 April 2007 19:54 |
Everything, no matter its size, weight, mass, speed, etc. comes in "shitloads". The only exception is "shit," which comes in "buttloads". |
http://www.cartalk.com/content/read-on/2000/03.10.html
Because someday you might need this information.
DC
Post by: Barry Hufker on April 03, 2007, 12:36:12 AM
Barry
Post by: Jørn Bonne on April 03, 2007, 09:22:25 AM
| Yannick Willox wrote on Sat, 31 March 2007 10:15 |
Nobody checked the graphs: > check out this website http://src.infinitewave.ca/.... and especially the r8brain PRO minimal phase against ALL of the others ? ??? |
Allow me to slide a bit OT here, but will the iZotope 64-bit SRC be available for the PC-platform anytime soon?
Also, is the r8brain PRO processing in 64-bit, like several of the other Voxengo plugs. Didn't see that mentioned on the website.
JB
Post by: Bob Boyd on April 03, 2007, 01:05:54 PM
| Matt_G wrote on Mon, 02 April 2007 09:45 | ||
It's funny Bob, I used to like soundBlade's current SRC (which isn't the iZotope at the moment). I still think that despite the aliasing it can sound great on certain material, can't explain the affect I hear but it kind of has a smooth soft sound which is great on some stuff. The iZotope SRC on it's high setting in WaveEditor can be a touch on the bright side but is likely the steepness of the default filter setting. See the attached image of the 'high quality' default setting in the current WE 1.2 version. With these being adjustable in 1.3, I bet we will get an amazing sounding SRC suitable for any style of music. With iZotope products it's all about high quality but giving the consumer all the options to create just what you need. MBIT+ is another example of having many options. Matt |
I'll be interested to see how the iZotope SRC sounds once the parameters are adjustable. It felt bright to me initially too.
Any idea what Sample Manager is using to dither the output file back to 24 bits? Is it MBit+?
I've been using pow-r on any SRC in Saracon.
Post by: Matthew Foust on April 03, 2007, 05:23:30 PM
| Ralf Kleemann wrote on Sun, 01 April 2007 05:27 |
What I also like about the company is that they are open about forthcoming releases and features. At Apple, you would already have been sacked for posting a screenshot to a forum... |
He'll be hearing from our lawyers. (Just kidding)
I hope I am not out of place responding on this forum. If so, my apologies.
Wave Editor 1.3 will bring a couple of other improvements with regard to iZotope's 64-bit SRC:
1) it will be multichannel (>2)
2) it will have 64-bit input
Post by: Matt_G on April 04, 2007, 08:44:25 AM
| Matthew Foust wrote on Wed, 04 April 2007 07:23 |
He'll be hearing from our lawyers. (Just kidding) |
Ahhh! good to have you here Matt
| Quote: |
I hope I am not out of place responding on this forum. If so, my apologies. Wave Editor 1.3 will bring a couple of other improvements with regard to iZotope's 64-bit SRC: 1) it will be multichannel (>2) 2) it will have 64-bit input |
Just to clarify what Matt has said further, at the moment the iZotope SRC implementation in 1.2 has 32bit input & 64bit internal DSP within the SRC. Whereas 1.3 will accept 64bit on the input as well.
Matt
Post by: Andy Krehm on April 10, 2007, 11:14:05 AM
| Bob Boyd wrote on Thu, 29 March 2007 01:30 | ||
I had this happen when I first tried Saracon too. You need to increase the buffers in the Preferences panel. The manual mentions this but may not stress it's importance. The optimal settings will vary from computer to computer but try opening the first buffer all the way and the second one about 3/4. (I'm away from my studio computer so I don't remember what the faders are labeled.) On my Dual 2.7 G5, a STEREO file going from 88.2k to 44.1/24 with pow-r will (at it slowest!) be over 2x realtime. Expect much faster results on the whole. |
Hey Bob:
I tried your suggestion and Saracon is lightning fast now. Thanks for the tip.
I'm sure this doesn't affect the quality of conversion or does it? Do you know?
Post by: Bob Boyd on April 10, 2007, 09:39:47 PM
Post by: Andy Krehm on April 10, 2007, 10:40:47 PM
| Bob Boyd wrote on Tue, 10 April 2007 21:39 |
Nope. Just optimizes the processing. Better, huh? |
Ah...sooo much better