Page 1 of 4

Western Digital Caviar SE16 500GB: Big Low Noise Champ?

Posted: Sat Jun 17, 2006 1:49 pm
by Devonavar

Posted: Sat Jun 17, 2006 3:25 pm
by theyangster
wow :shock: that's very impressive

screw solid state! :lol:

j/k

Posted: Sat Jun 17, 2006 3:28 pm
by Lards
Yay! Finally a decent sized hard drive with low noise! :)

As good as my Samsung 1614C's are, 160gb is getting a little small - and when you have 6 of them the vibration is a big issue. So even though vibration is similar with the WD it should be a lot less of a problem with only 1 or 2 of them.

Posted: Sat Jun 17, 2006 4:54 pm
by Ackelind
Nice!

Finally a drive with both capacity and low noise. Sure hope this is a trend that others will follow.

Posted: Sat Jun 17, 2006 5:28 pm
by CA_Steve
Nice!

Found a typo. At the bottom of page 2, this sentence's URL "Many thanks to Western Digital for the Caviar SE16 sample" points to seagate.com :shock:

Posted: Sat Jun 17, 2006 5:58 pm
by SpHeRe31459
I guess now the question would be to be: how does one make sure you get a quiet model aside from physical inspection? Since, as was mentioned in the article, WD likes to change the drive characteristics and not change the model number.

Posted: Sat Jun 17, 2006 7:57 pm
by winguy
Anyone can point me to the official platter information for the SE16 product line? Only 250GB and 320GB show 3 platters.

Posted: Sat Jun 17, 2006 9:07 pm
by qviri
Wow... Kind of makes me uncertain about the validity of my decision to buy a P120 last week.

Let's hope Samsung react to the challenge. Is there any chance of having SPCR review the new-ish T133? I saw a 300 GB model when I was buying my 250 GB, and kind of wondered how well that would work.

Posted: Sat Jun 17, 2006 10:26 pm
by spolitta
impresive results,

Posted: Sun Jun 18, 2006 12:23 am
by rpsgc
Wow, that's a nice surprise! I always had a feeling these new black HDDs were somewhat "quiet" but never expected this :)

Posted: Sun Jun 18, 2006 2:09 am
by smilingcrow
The WD Cavier SE specs claim that it 2 dB quieter than the Cavier SE16 across the board; it has a maximum capacity of 320GB and an 8MB cache. On paper that comes across as the quietest desktop drive available; it would be nice to see a review of that.

Cavier SE
Idle Mode 26 dBA (average)
Seek Mode 0 31 dBA (average)
Seek Mode 3 27 dBA (average)

Cavier SE16
Idle Mode 28 dBA (average)
Seek Mode 0 33 dBA (average)
Seek Mode 3 29 dBA (average)
qviri wrote:Let's hope Samsung react to the challenge. Is there any chance of having SPCR review the new-ish T133? I saw a 300 GB model when I was buying my 250 GB, and kind of wondered how well that would work.
This Review shows that the T133 performance is very significantly below that of the P120. They were impressed with noise and heat levels, but there was no objective noise data from what I can remember. Maybe the 400GB will address this issue. Is it released yet?

Posted: Sun Jun 18, 2006 2:32 am
by Shining Arcanine
The Caviar SE and the Caviar SE16 are pretty much the same thing but with different caches. The reduction in noise in the 320GB Caviar SE should come from having three platters instead of four, meaning that you should get the same reduction from the SE16 version.

Posted: Sun Jun 18, 2006 3:49 am
by Ackelind
This drive looks like a better choise than Samsungs's T133 if you need this amount of storage.

This drive is on par with the P120 (even better regarding seeks), and the T133 is reported so (at least subjectively) be louder than the P120.

Posted: Sun Jun 18, 2006 4:20 am
by smilingcrow
Shining Arcanine wrote:The Caviar SE and the Caviar SE16 are pretty much the same thing but with different caches. The reduction in noise in the 320GB Caviar SE should come from having three platters instead of four, meaning that you should get the same reduction from the SE16 version.
WD quote the same noise figures for the whole range of SE16 KS models (250 – 500GB) and they do the same for the SE JS models (160 – 320GB); which leaves us all guessing as to how they compare. They don’t seem to list platter counts either, which doesn’t help.

This comparison shows the WD2500KS performing somewhere in between a P80 & P120. It also shows the noise level to be much higher than the Samsungs; possibly a different revision!
I’m wondering how the SE JS compares to the SE16 KS in the performance stakes; 8MB of cache versus 16MB shouldn’t make a big deal surely? There may be other differences that affect performance of course though.
I have a choice of two WD3200JS for £140 or one WD5000KS for £160. I’d prefer the two WD3200JS drives provided they don’t perform as poorly as the T133. I don’t mind a small performance hit , but the T133 is just too slow.
Ackelind wrote:This drive looks like a better choise than Samsungs's T133 if you need this amount of storage.
This drive is on par with the P120 (even better regarding seeks), and the T133 is reported so (at least subjectively) be louder than the P120.
Samsung’s figures back that up with the P120 being 2.5/2.8 Bel versus the T133 2.7/2.9 Bel; this is for idle/random read/write.

Posted: Sun Jun 18, 2006 7:35 am
by MikeC
The caution I make here is the same one I make on most noise-making products we review: Sample Variance is very real, especially in acoustic characteristics because they tend to be less tightly controlled than many other performance aspects in most products. It will be interesting to run a poll in the forums about this noise of this model as perceived by owners. Once we have a sampling of at least 10, preferably 20, then we can be more confident that our exceptionally quiet sample is not an anomaly.

FYI, I believe these factors affect the acoustics of each & every HDD drive:

1) sample variance -- some just leave the factory a little noisier or quieter, sometimes in only one state (like idle, seek, write), sometimes in all/most states. My bet is that if they were sorted at end of the production line by noise, you'd have at least 2 classes for each model: normal and quieter. But then a 3rd category seems likely: Noisier. By binning them in this way, drive makers could probably charge a premium for the quieter variants.

2) subtle damage in transit -- drives are shipped at least halfway around the globe in almost every instance. Yes, there's great care taken to pack them well, but surely -- like fans -- some might be subtly changed by external shock or exposure to temperature / humidity extremes. Not enough to affect performance in any serious way, but enough to affect the noise. (You know, how much do you think it would take to increase the noise a dB or two?)

Posted: Sun Jun 18, 2006 9:33 am
by Devonavar
SpHeRe31459 wrote:I guess now the question would be to be: how does one make sure you get a quiet model aside from physical inspection? Since, as was mentioned in the article, WD likes to change the drive characteristics and not change the model number.
You have to look at the revision number that is appended to the model number on the label. Resellers should be able to give you this information.
This review examines the most recent revision of the Caviar SE16 series, coded 00MNB0 as a suffix to the model number.

Posted: Sun Jun 18, 2006 9:36 am
by Devonavar
winguy wrote:Anyone can point me to the official platter information for the SE16 product line? Only 250GB and 320GB show 3 platters.
I couldn't find any. My source of information was a news blurb on StorageReview that it has four 125GB platters. Given that storage review has long-term contacts with just about every drive manufacturer, I consider them a reliable source.

Posted: Sun Jun 18, 2006 12:32 pm
by ~El~Jefe~
wow nice find.

finally i can get a real hardrive. 160 gb spinpoint 80 is a nice thing to have, but not forever. I guess this drive demands suspending as it has more vibration than other drives.

16 meg cache and 4 platters at a much higher density. that's probably a fast drive.

Posted: Sun Jun 18, 2006 6:46 pm
by JazzJackRabbit
The noise numbers look very nice. However, when listening to sound samples P80 sounds much much better. WD noise is much harsher and higher toned. Unless there was a diffence in the way recordings were made ie distance, angle, sound volume, I would never chose WD over P80/P120.

Can Devon or anyone else from SPCR crew explain why P80/WD noise measurements are almost the same, but WD sounds so much worse than P80?

Posted: Sun Jun 18, 2006 8:29 pm
by Woland
I come here, see the article about the black P150 and think, "The only thing that stopped me from picking one up and building a computer around it was that it was white." Now that it's available in black, it's exactly the case I want. And I wouldn't have heard about it if not for SPCR.

Then I see this article about a 500 GB drive that doesn't rattle like a freight train and think, "Damn that does it. A P150 and this drive are it! Time to build a new box."

We bought a bunch of Seagate 500 GB drives for a big huge RAID array at about $380 each when they were first released. They're fine drives and all, but to be able to get the same capacity for about a grand less just 3 months later (and that produce much less vibration to boot)...God I love the relentless pace of computing technology.

Antec and WD should totally float a couple bucks SPCR's way for this kind of exposure :wink:

Posted: Sun Jun 18, 2006 9:33 pm
by MikeC
JazzJackRabbit wrote:The noise numbers look very nice. However, when listening to sound samples P80 sounds much much better. WD noise is much harsher and higher toned. Unless there was a diffence in the way recordings were made ie distance, angle, sound volume, I would never chose WD over P80/P120.

Can Devon or anyone else from SPCR crew explain why P80/WD noise measurements are almost the same, but WD sounds so much worse than P80?
You have to remember that the recording mic is 3" away. We do this because if we put it 1 meter away (where we take the noise measurement), you'd never hear half the stuff we record over the noise floor of your audio playback gear -- and our ambient noise. So whatever difference there is gets exaggerated. I think I can agree that in real life, from a few inches away, the Samsung sounds nicer, but from the typical 1/2 to 1m distance and with the HDD floating in an elastic suspension inside a sturdy case, I'd never hear that difference.

Posted: Mon Jun 19, 2006 1:27 am
by cAPSLOCK
It's performance has been reviewed at Hardware.fr, apparently decent performance.

One worry is that they didn't find it to be exceptionally silent (as you would expect from a drive that has dethroned the P80), they said it was "relatively discrete, but not silent" and found it comparable to the seagates. I know they are far from being experts in acoustics like SPCR, but it does reinforce the concerns of sample variance...

On another note, where I live, it's very expensive, you can buy three P120 250GB for a bit less than the WD drive costs.

Posted: Mon Jun 19, 2006 4:28 am
by MikeC
cAPSLOCK wrote:It's performance has been reviewed at Hardware.fr, apparently decent performance.

One worry is that they didn't find it to be exceptionally silent (as you would expect from a drive that has dethroned the P80), they said it was "relatively discrete, but not silent" and found it comparable to the seagates. I know they are far from being experts in acoustics like SPCR, but it does reinforce the concerns of sample variance...

On another note, where I live, it's very expensive, you can buy three P120 250GB for a bit less than the WD drive costs.
Consider the distance of Hardware.fr's recording microphone: 5cm from the HDD. This is a serious problem. There's no way the decibel reading can be accurate due to boundary effects. It's the same problem at storagereview.com -- not even relative differences are necessarily correct due to compression effects; the close proximity impacts every measurement similarly, reducing differences. From the rough babelfish translation, I get the distinct impression that they had a casual, passing listen; they did not really study the acoustics, either subjectively or objectively.
A test of hard disk and their function AAM would be incomplete without a measurement of the sound harmful effect of the discs, measured here to approximately 5 cm. This would not be that for the noise in rotation it y' has notable differences. Thus, if the discs Western and Seagate are relatively discrete, without being quiet, it is not at all the case of the models Hitachi and Maxtor. Those are caught up with because of a less overcost at the time of the access, but one remains all the same beyond what proposes the Western disc. To change the Seagate discs are noisiest when one reaches in an intensive way the disc, and the AAM can nothing there since Seagate does not propose this function.

Posted: Mon Jun 19, 2006 6:01 am
by JazzJackRabbit
MikeC wrote: You have to remember that the recording mic is 3" away. We do this because if we put it 1 meter away (where we take the noise measurement), you'd never hear half the stuff we record over the noise floor of your audio playback gear -- and our ambient noise. So whatever difference there is gets exaggerated. I think I can agree that in real life, from a few inches away, the Samsung sounds nicer, but from the typical 1/2 to 1m distance and with the HDD floating in an elastic suspension inside a sturdy case, I'd never hear that difference.
OK, I see. Recording sound from 3" away will certainly exaggerate any differences there are. However in my experience I can hear the difference between different drives even 0.5-1 meter away. Any difference in quantitative noise will fade away as you move further away but it still will be there and most importantly the noise character will stay the same. In other words the harsher character of WD drive will be heard even from one meter away. I don't know if it will be the same if the drive was put inside the case, but I would imagine under the strictest test conditions I would hear the difference. The argument about playback gear or ambient noise is useless as it varies too much. Right now, in the summer with the windows open I can't hear my PC even with all fans cranked up to 12V (due to summer heat), however during winter with windows closed, no airconditining noise and no birds chirping outside the window (stupid fluffy creatures :rolleyes: ), I can hear my PC, I can hear my 2504C neatly tucked away inside the case as far as possible and I can hear my fans if they rotate above 800 or so RPM. Oh well, that's just my opinion.

Posted: Mon Jun 19, 2006 6:55 am
by MikeC
JazzJackRabbit wrote:Any difference in quantitative noise will fade away as you move further away but it still will be there and most importantly the noise character will stay the same. In other words the harsher character of WD drive will be heard even from one meter away.
I'm not sure I agree with you. When the overall noise level drops because of distance, it becomes harder to hear the quality differences of the noise, too. The recordings are not 100% true representations of what we hear in the lab -- they definitely tend to exaggerate the differences. When it is heard from a meter away, or with the PC under the desk, the character differences at idle becomes very subtle, and the more apparent difference is the lower noise floor at seek -- the WD is quieter than any other drive in seek by at least 2 dBA@1m.

An Aside: Maybe we need to look at increasing the distance between mic and noise source -- to perhaps a foot or half a meter, which would still give us some headroom above the ambient noise floor (of the room and the audio electronics). It might make the recordings sound a bit more realistic.

My original intent in placing the mic so close was to give everyone the effect/benefit of putting one's ear next to the component the way I/we often do in the lab to examine the noise. It also eliminates the issue of background or electronics noise interference, because the foreground noise completely dominates. But it is not a realistic representation of the noise from a more nominal distance that users would have, partly due to the excess of detail and boundary-effect emphasis in lower frequencies. The trouble with changing the recording distance is that all the previous SPCR recordings of the last 2+ years would no longer be directly comparable. We'll have to think about this and see if there is a solution. Perhaps we can judiciously add the longer distance recordings... at the expense of yet more time spent for each review. :roll:

Posted: Mon Jun 19, 2006 7:08 am
by jaganath
Any difference in quantitative noise will fade away as you move further away but it still will be there and most importantly the noise character will stay the same.
Surely that's not true, because the higher frequencies are more quickly attenuated by the atmosphere than lower ones. So at a greater distance the lower frequencies will predominate, and the noise "character" will be different.

I think measuring the sound levels from less than a metre away is fairly realistic; if I was to take a wild guess I would say the average "operator position" is about 0.5-0.75m from the PC (if only because most mouse and keyboard cables are not much longer than this, and yes I know people use wireless mice and kb).

Posted: Mon Jun 19, 2006 7:50 am
by MikeC
jaganath wrote:I think measuring the sound levels from less than a metre away is fairly realistic; if I was to take a wild guess I would say the average "operator position" is about 0.5-0.75m from the PC (if only because most mouse and keyboard cables are not much longer than this, and yes I know people use wireless mice and kb).
Measuring distance is not an issue, imo. I know ISO7779 specifies ~0.6m, but I prefer the 1m distance because...
1) so many official acoustics specs are done at this distance, which means our results can be directly compared.
2) the difference between 1m and 0.6m is perhaps 1 dBA, which is simple to adjust for.

My comments in the previous post was referring specifically about audio recording distance.

Posted: Mon Jun 19, 2006 8:08 am
by m_yates
sample variance -- some just leave the factory a little noisier or quieter, sometimes in only one state (like idle, seek, write), sometimes in all/most states.
So is there a chance that the one at SPCR was "hand picked" for review?

Posted: Mon Jun 19, 2006 8:25 am
by MikeC
m_yates wrote:So is there a chance that the one at SPCR was "hand picked" for review?
Judging by the packaging, it seems unlikely, but as mentioned before, the best way to confirm our measurements & reactions is a user poll. Surely some people here are alreadying using this drive, as it's been out for many months.

Posted: Mon Jun 19, 2006 8:27 am
by smilingcrow
m_yates wrote:
sample variance -- some just leave the factory a little noisier or quieter, sometimes in only one state (like idle, seek, write), sometimes in all/most states.
So is there a chance that the one at SPCR was "hand picked" for review?
Since the drive under review was supplied by Western Digital, it’s reasonable to assume that they made sure that a good sample was sent. As long as the sample was typical for this drive and not a one in twenty ‘golden’ sample, I don’t see a problem with this policy. If their policy is more about making sure they didn’t provide a rogue drive, in the acoustic sense, then that seems reasonable to me. Anything more would be plain misrepresentation.