Pointers to good AD

Posted: Sun Aug 28, 2005 8:09 pm

Good points Alex!!

I guess I should be more cautious in my recommendations as the converters I recommended and the comments I made above are in fact based on subjectivity. In otherwords - you get what you pay for and you just got some *free* advice from me Gunnar... Laughing

J.

Posted: Sun Aug 28, 2005 9:47 pm

Cucco wrote:

Good points Alex!!

I guess I should be more cautious in my recommendations as the converters I recommended and the comments I made above are in fact based on subjectivity. In otherwords - you get what you pay for and you just got some *free* advice from me Gunnar... Laughing

J.

Ahh - maybe this one is flying over my head...what is the *free* advice?

Alex

Posted: Mon Aug 29, 2005 1:42 am

Cucco wrote:

In otherwords - you get what you pay for and you just got some *free* advice from me Gunnar... Laughing

J.

I think I have received a lot of really good advice on this forum from sincere people sharing their both objective and subjective experiences! And free as well. The rest as always is up to me.

I am in the process of looking through my computer setup as my old laptop is showing signs of age and has always been noisy.. If I go with a new laptop, a firewire connection is probably a must from the sound card (maybe the Lynx Aurora with the promised FW option), or perhaps a "digital" to FW card + a separate converter.

Well, to keep it short, I am learning a lot here. The last word is not said yet.

Gunnar

Posted: Mon Aug 29, 2005 3:39 am

alexaudio wrote:

Cucco wrote:

Good points Alex!!

I guess I should be more cautious in my recommendations as the converters I recommended and the comments I made above are in fact based on subjectivity. In otherwords - you get what you pay for and you just got some *free* advice from me Gunnar... Laughing

J.

Ahh - maybe this one is flying over my head...what is the *free* advice?

Alex

My advice re: the converters = *Free*
Wink

IOW - YMMV, etc...

Posted: Tue Aug 30, 2005 10:13 am

http://www.lake-people.de/

Check out these converters. There was a test on the net where the authors compared the Lavry Gold, StageTec, Prism, Lake People and some other hi end A/Ds. While the StageTec came out on top, the Lake People converter was a really close second, leaving the other converters further behind. They have a 4 ch A/D converter for 24-25000 skr and are distributed by polteknik (www.polteknik.se).

For the record, I´ve heard a StageTec console at Berwaldhallen in Stockholm and it sounded... good. If the Lake People converter isn´t far behind, it might be worth listening!

Mats

Posted: Tue Aug 30, 2005 12:43 pm

route909 wrote:

http://www.lake-people.de/

Check out these converters. There was a test on the net where the authors compared the Lavry Gold, StageTec, Prism, Lake People and some other hi end A/Ds. While the StageTec came out on top, the Lake People converter was a really close second, leaving the other converters further behind. They have a 4 ch A/D converter for 24-25000 skr and are distributed by polteknik (www.polteknik.se).

For the record, I´ve heard a StageTec console at Berwaldhallen in Stockholm and it sounded... good. If the Lake People converter isn´t far behind, it might be worth listening!

Mats

Thanks for the information. I took a look at the Lake People ADCs sometime ago and found that they do not support 88.2 kHz sample rate. This became a limitation, as much of the work here for my clients is done at the sample rate.

Posted: Tue Aug 30, 2005 1:44 pm

That's a pretty odd reason to shoot down a converter?? Would not the clients be fine with a 96 kHz rate?

Posted: Tue Aug 30, 2005 2:44 pm

alexaudio wrote:

Thanks for the information. I took a look at the Lake People ADCs sometime ago and found that they do not support 88.2 kHz sample rate. This became a limitation, as much of the work here for my clients is done at the sample rate.

That´s odd. This page says it´s 88.2 kHz compatible (32-192 kHz): http://www.lake-people.de/Digital_devices_Deutsch/F444de/f444de.html

Mats

Posted: Tue Aug 30, 2005 3:39 pm

Cucco wrote:

That's a pretty odd reason to shoot down a converter?? Would not the clients be fine with a 96 kHz rate?

The scenario required multiples of 44.1, which was required when I was looking. Not trying to shoot down anything, my appologies if it came across that way. Looks like they do have a nice unit, one which wasn't around when I was looking.

Seems the new one goes up to 192kHz. Has anyone read Lavry's documentation on not needing to utilize sampling rates higher than 96khz? Just curious how many people have read his papers and information.

Posted: Tue Aug 30, 2005 4:22 pm

alexaudio wrote:

Cucco wrote:

That's a pretty odd reason to shoot down a converter?? Would not the clients be fine with a 96 kHz rate?

The scenario required multiples of 44.1, which was required when I was looking. Not trying to shoot down anything, my appologies if it came across that way. Looks like they do have a nice unit, one which wasn't around when I was looking.

Seems the new one goes up to 192kHz. Has anyone read Lavry's documentation on not needing to utilize sampling rates higher than 96khz? Just curious how many people have read his papers and information.

I bet a lot of people here have, it´s quite a knowledgeable forum. My post merely pointed out it´s compatibility with 88k. I have an RME Fireface 800 and haven´t even tried recording at 192, it´s a bit of a waste i m o.

BTW, have you read Bob Katz book on mastering? He states that dynamic processors sound better at higher sample rates and proves it with some graphs. If I´m not entirely wrong, eq is supposed to sound better too. My ears tell me that anyway. I sometimes record at 96 when making choir stuff or solo pieces with small track counts. Though for most of my work 44.1 is good enough.

Do you do much hi res work?

Mats

Posted: Tue Aug 30, 2005 5:04 pm

route909 wrote:

alexaudio wrote:

The scenario required multiples of 44.1, which was required when I was looking. Not trying to shoot down anything, my appologies if it came across that way. Looks like they do have a nice unit, one which wasn't around when I was looking.

Seems the new one goes up to 192kHz. Has anyone read Lavry's documentation on not needing to utilize sampling rates higher than 96khz? Just curious how many people have read his papers and information.

I bet a lot of people here have, it´s quite a knowledgeable forum. My post merely pointed out it´s compatibility with 88k. I have an RME Fireface 800 and haven´t even tried recording at 192, it´s a bit of a waste i m o.

BTW, have you read Bob Katz book on mastering? He states that dynamic processors sound better at higher sample rates and proves it with some graphs. If I´m not entirely wrong, eq is supposed to sound better too. My ears tell me that anyway. I sometimes record at 96 when making choir stuff or solo pieces with small track counts. Though for most of my work 44.1 is good enough.

Do you do much hi res work?

Mats

Yeppers - read Bob's book cover to cover...highly recommended read. The majority of the work I do is now in high resolution, usually 88.2 kHz or 96kHz. Much of it is now in surround as well. Far as dynamics and EQ sounding better at higher sampling rates, I find this to be true, so long as the application/hardware is specifically written and supports the higher sampling rates. Far as EQ sounding better, this makes sense, due to the filter sets involved. However, one needs to look at all variables. If the finished product is intended for release on CD, the EQ may sound better at a higher sampling rate, but is the difference noticable after SRC. Then comes the variable of the quality of SRC. If I know the the end result is going to be on CD and if the project warrants it, I record at 88.2kHz so the SRC is a more simplistic 2 to 1 conversion.

Far as utilizing 192kHz, I have found no significant advantage of that sample rate. I find Lavry's papers a very interesting and provocative read, as well as the statements of utilizing converters above 96kHz providing diminishing return in some regards to be valid.

Time for a topic switch?

Posted: Wed Aug 31, 2005 12:52 am

alexaudio wrote:

Then comes the variable of the quality of SRC. If I know the the end result is going to be on CD and if the project warrants it, I record at 88.2kHz so the SRC is a more simplistic 2 to 1 conversion.

Alex-

I see this argument brought up a lot and I've seen many arguments about SRC quality versus processing quality. I believe it was Daniel Weiss wrote a white paper some time back about how the difference was basically negligible when resampling from 88.2 versus 96K. The thought process basically being that when we resample, it isn't just a removal of half of the samples- as you would be loosing much data. Also added to that the fact that there are some pretty complex filters used to deal with anti-aliasing. Between those two, the math ends up being pretty much a moot point.

What are your feelings about this? I do all of my work at 96K. To be honest, I don't hear a huge difference between 88.2 and 96 but I just opt to process at the higher rate.

--Ben

Posted: Wed Aug 31, 2005 2:30 pm

FifthCircle wrote:

alexaudio wrote:

Then comes the variable of the quality of SRC. If I know the the end result is going to be on CD and if the project warrants it, I record at 88.2kHz so the SRC is a more simplistic 2 to 1 conversion.

Alex-

I see this argument brought up a lot and I've seen many arguments about SRC quality versus processing quality. I believe it was Daniel Weiss wrote a white paper some time back about how the difference was basically negligible when resampling from 88.2 versus 96K. The thought process basically being that when we resample, it isn't just a removal of half of the samples- as you would be loosing much data. Also added to that the fact that there are some pretty complex filters used to deal with anti-aliasing. Between those two, the math ends up being pretty much a moot point.

What are your feelings about this? I do all of my work at 96K. To be honest, I don't hear a huge difference between 88.2 and 96 but I just opt to process at the higher rate.

--Ben

I have not read Daniel's paper, but certainly would love to. I'll email him when I have a chance to get a copy if available. I have just made it a general rule of thumb of, if I know it will be ending up on CD, I just start at 88.2. If I know it will be on DVD-A - it is a crap shoot really. If it is DVD-V or crossbreed - I will indeed start with 96. Either way, I think the results between 88.2 and 96, stealing your words, are pretty much moot - especially when using quality software SRCs available. The only case where I notice the difference, which is small, is when I have to use a realtime SRC hardware unit which isn't of the highest quality. In that case, I do notice a, be in small, difference with 88.2 vs. 96 SRC'ing to 44.1. In that case, 88.2 to 44.1 in realtime seems to resemble more the original in soundstage depth. That is about as specific as I can get, not really scientific either, as I haven't performed a true AB test. I am sure that if I had a Weiss unit, it wouldn't matter at all.

I find that maintaining overall system integrity, clocking and 24 bit depth is more important that the 88.2 vs. 96 - don't you?

Alex

Posted: Wed Aug 31, 2005 5:48 pm

alexaudio wrote:

I find that maintaining overall system integrity, clocking and 24 bit depth is more important that the 88.2 vs. 96 - don't you?

Alex

Absolutely... I found that 24 bit recording gave me by far the greatest increase in fidelity of my recordings (well, that and switching to recording multitrack in the DAW instead of mixing outboard and going straight to stereo, but that's another topic altogether).

If you write to Daniel Weiss, I'd love to hear his current thoughts. It was quite awhile ago that I read the paper (and as I said before, I believe he was the author, but I could be mistaken).

--Ben

Posted: Fri Sep 02, 2005 4:47 pm

alexaudio wrote:

Then comes the variable of the quality of SRC. If I know the the end result is going to be on CD and if the project warrants it, I record at 88.2kHz so the SRC is a more simplistic 2 to 1 conversion.

FifthCircle wrote:

I see this argument brought up a lot and I've seen many arguments about SRC quality versus processing quality. I believe it was Daniel Weiss wrote a white paper some time back about how the difference was basically negligible when resampling from 88.2 versus 96K. The thought process basically being that when we resample, it isn't just a removal of half of the samples- as you would be loosing much data. Also added to that the fact that there are some pretty complex filters used to deal with anti-aliasing. Between those two, the math ends up being pretty much a moot point.

I've written multi-rate DSP code, including polyphase SRC filters. In theory I agree with Daniel. In practice, I don't really feel safe trusting a randomly selected DAW's non-integer-ratio conversion routine to be written correctly. Even though the (long) prototype filter is very similar in either case, the associated program logic that decides which filter taps to use in which phases, and which phases happen when, is a lot more complex. If it's done right, then no problem. If it's done wrong... Evil or Very Mad

Ben and Alex, what do you hear in the Samplitude SRC routines? I just converted a surround project from 88.2 to 48 in order to put it on a DVD-R using Disk Welder Steel, and I wasn't altogether pleased with the results. I took the wizard's suggestion of using the not-quite-slowest routine because I was dealing with 24-bit files. But I'm sure they were 32-bit data by the time they hit the SRC routine, so I need to try the best option offered and see if I like it any better. Maybe it's just that I don't like 48k as much as I like 88.2. Whatever the problem was, it sure dried things up.

David L. Rick

This topic sponsored by: