Level 1

15 points

Mini Mac crunches IMac G5! I'm seriously deluded!

I have a singular story to tell you. Maybe you have some good advice.
I am very surprised about what follows:

A couple of months ago I stepped into MacMondo getting myself a MINI G4, 1.42GHz and upgraded ram to 512 MB. HD is 80GB/4200rpm. Old IBook standards, right?
Today I took my second step into Apple, buying a gorgeous brand new IMac G5 20"display, 2.1GHz cpu and upgraded to of 1.5GB ram. HD is 250GB SATA/7200rpm.

As anybody would expect, the G5 should massively crunch the Mini, right? Wouldn't you?

To my sad surprise here's what happened:

Took a DVD I own and used MacTheRipper to import it to HD.
Mac Mini : took 27 min and 18 sec
IMac : took 32 min and 46 sec

Once the files were on both HDs, I encoded them in MP4, 1650 kbps/720 x 384 size/ audio @44.1KHz 128 kbps...using Hand Brake.
Both machies encoded an MP4 of about 1.3x GB. The MINI was definitely faster, again!

Last test: CD ripping & MP3 encoding:
inserted Marc Johnson latest CD (I own it) in each mac, using the native ITunes. Both machines run OS 10.4.3 and are identically updated to latest Apple software.
Well...the CD lasts exactly 1 hour.
The encoding was to MP3 at 192kbps...the top quality allowed by ITunes.

Here we go again:
MINI : took 3 min and 46 sec
IMac : took 9 min and 57 sec

In all these 3 tests I monitored the Monitor Activities app, checking cpu and ram loads.... the Mini consistently showed less loads than the IMac!!!

????????

This is ridiculous!
Much faster and latest generation cpu/ faster HD/ bigger and faster RAM/same OS and apps..... worst performance!
Not to mention poorer video quality...even though the Imac sports an ATI X600 w/128MB ram on PCI-Xpress....against the MINI ATI 9200 w/32MB ram, previous generation graphic cpu.
The machine is "virgin"...opened the sealed box!

Should I think the G4 is better unless one buys a dual G5 Mac?
Which is want I don't intend to do anyway....it's ridiculous....

BTW, my 2.8GHz Pentium4 kills both units by long measure in these 3 apps.

Maybe I should go back and get a PowerBook 15" w/G4 1.67GHz and 1 GB of Ram...and that'll work at least as good as the Mini if not better.

Any thoughts, folks?

Very appreciated...

best
Italo

Edited by an Apple Host

Imac 20", Mac OS X (10.4.3)

Posted on Nov 22, 2005 4:10 PM

61 replies

Italoop Author

Level 1

15 points

Nov 23, 2005 7:53 AM in response to Rod Hagen

Hi Rod

first...let me thank you for be so supportive and encouraging: I really appreciate it.
Here are some more interesting details after evolution of the situation.
This morning called my Mac dealer and reported the problem.
Went downtown to do more tests on different Macs in the store. An expert was on location for good support.
We tested the following:
my 1.42/512MB ram Mini Mac
an Isight IMac 17" G5, 1.9GHz/512MB ram
a new 1.67GHz/512MB ram G4 PowerBook
a dual 2.7GHz/3.5 GB ram G5 Power PC (monster!)

The 2 store folks (they are expert!) remained speechless.
Turns out that the my Mini rips/encodes to MP3 faster than my Imac 20" and the location IMac 17" (both ISight). The PowerBook is in between, sligtly slower than my Mac Mini, much faster than the 2 Imacs.
The Power PC runs faster than everything...showing the same exact performance I get from a Fujitsu-Siemens P4 3.2GHz 775 socket/1GB ram...which costs 4 time less than the Power PC. (???)
Ripping legally a DVD I own and want to have HD copy of it for my leisure...
The Mac Mini beats both IMac Isight 17" and 20".
The Power Book is again in between the Mini and the IMacs.
The Power PC runs faster than everything...at same performance level and speed as my already mentioned Win PC.
Encoding the 6.5GB VOBs into MP4:
Mac Mini slighlty beats both IMacs. Power Book is in the same league as the Mini.
Power PC runs fast! 43 minutes....same as my 3.2GHz pc.

The only test where the Imac proved to be faster than the Mini:
a cd was ripped in AIFF format to HD. After this, the files were encoded to MP3, using ITunes. Please note the difference between the first test.
In the first one, audio files were ripped AND encoded to MP3 at the same time, from the cd. In this last tests MP3s were encoded from files already stored to the HD.
The IMac run faster than the Mini and the PowerBook....
We were able to see that the Imac disk driver (CD/DVD reader-burner) is actually slower than the Mini, PowerBook and Power PC drives.
That is the reasons for lower performance on "from disk" acquisition/encoding. The hardware doesn't do justice to the IMac.
Nevertheless the video encoding from hd stored VOB files to MP4 (so...no external disk involved in this) turned out to be..STILL...better on the Mini and PowerBook than the IMacs.

What we all learned:
my IMac has no issues. IMacs work like this.
Mac MINI and PowerBook (G4 dude) are very good tools for these apps.
Power PC...that's the only machine that can be compared to what a pretty standard pc can do today, with similar tasks/apps.

Rod:

both IMacs work at pretty much same level. Don't need days to get fast.
Monitoring apps for CPU hog: as already reported nothing is taking anything from the G5 unit. All open apps stay on 0%,,,,some hit 2 or 3% sometimes.
The used apps are fully compatible w/Tiger.
All Macs in this tests run 10.4.3.
There is nothing in a big dvd that should keep the G5 at lower performance than a G4, nor legal issues 8which I don't understand why you mention this...as a legal owner can legally make his/her own copy anyway).
I haven't run other benchmarks...but I don't see why I should.
I expect that other apps might show a better performance on the G5 IMac...I would HOPE so. The point is that 3 pretty mainstream apps like encoding MP3s, ripping and encoding Video...something most people do on any computer of any brand/OS....turned out to be significatively poor on the IMacs...and better on lower G4s.
The monster Dual 2.7GHz G5 powerPC/3.5GB ram behaves exactly like a P4 3,2GHz...it makes no sense to cash out 4 times the money for these tasks, obviously.

Thoughts:
-maybe I should just get a PowerBook and be happy with "portatility".
It works pretty good for my software job and for leisure.
-maybe I could get a Dual 2GHz G5...staying pretty much in the same money area than the Imac 20" Isight.....maybe?
-maybe it would be smarter to wait for Intel Macs. those I'm sure...will be fast enough to justify all this hype we keep hearing.....

Suggestions?

Thanks
Italo

Nov 23, 2005 12:21 PM in response to Italoop

We were able to see that the Imac disk driver (CD/DVD reader-burner) is actually slower than the Mini, PowerBook and Power PC drives.

That is the reasons for lower performance on "from disk" acquisition/encoding. The hardware doesn't do justice to the IMac.

The Superdrive in the iMac is actually faster than the one in the Mini. The problem is that some manufacturers limit the rip speed of commercial CDs and DVDs on their drives with the firmware. Why? According to the manufacturers that do this is to prevent the use of their drives for piracy.

As an example Pioneer on its DVR-109 superdrive model limits the ripping speed of commercial DVDs to 4x but the drive is capable of 12x which you can achieve by flashing the drive with an "unofficial" firmware. So how long it takes to rip a commercial CD or DVD to your HD is not indicative of how powerful a system is.

The point is that 3 pretty mainstream apps like encoding MP3s, ripping and encoding Video...something most people do on any computer of any brand/OS....turned out to be significatively poor on the IMacs...and better on lower G4s

There is no way that encoding MP3s or video is slower in the Imac that in the Mini. Did you set the iMac processor performance to highest?

Italoop Author

Level 1

15 points

Nov 23, 2005 12:48 PM in response to Rafael Marino

I agree.
There is a good possibility the CD/DVD drive is simply intentionally limited in its performance. As I stated in my previous posting, if I work on a file that is already stored to the Imac HD, the performance is generally faster than on the MINI, particularly audio. On Video files the Mini and the IMac seem to be on the same level. Don't understand why, really.
As more tests I tried to do some Wav to FLAC encoding:
Mini took 14m:25sec to encode 4 long wav files from the HD.
IMac took 11m:20sec to do the same job. Better...but not impressive!

Encoding a 26 minutes long/44.1KHz_16bit WAV file to MP3 @ 192 kbps took:
MINI : 3m:5sec
IMac : 1m:45sec
that's a better performance!

Yes, the cpu is set to MAXIMUM.

thanks
Italo

Nov 23, 2005 1:22 PM in response to Italoop

As more tests I tried to do some Wav to FLAC encoding:

Mini took 14m:25sec to encode 4 long wav files from the HD.
IMac took 11m:20sec to do the same job. Better...but not impressive!

Remember that not all applications are perfectly optimised. You can install Cee Pee You and it will show in the menu bar the CPU load so you can easily see when an application is actually using all the cycles that are available.

Rod Hagen

Level 7

32,000 points

Nov 23, 2005 1:40 PM in response to Italoop

nor legal issues 8which I don't understand why you mention this...as a legal owner can legally make his/her own copy anyway).

Sorry Italo, this actually varies greatly from country to country, and I forgot you were in Italy. Its not legal in Australia, even as a backup (though many people still do it, of course). In the US, where most people here come from, as I understand it, even the relevant software is illegal.

I'll post back on the technical issues later. Cheers

Rod

Rod Hagen

Level 7

32,000 points

Nov 23, 2005 5:14 PM in response to Italoop

Still not sure what speed you were running the iMac at teh store in, during these tests, Italo.

Have a look at http://www.macintouch.com/perfpack/comparison.html for an example of the quite dramatic differnce in many tasks between using "automatic" and "highest" processor speed on the iMac. As you will see, the mini does, in fact, outperform the iMac when the iMac is in "automatic" mode, but is well and truly walloped when the iMac is set to "highest".

Drop me an email for some thoughts on the DVd issues - my address is in my profile.

Cheers

Rod

Kerr

Level 1

35 points

Nov 23, 2005 8:23 PM in response to Italoop

I may have the answer for mac the ripper. If you are not using the version 3 beta (I don't think three is fully out yet) then you are lucky that MTR is working on the g5 at all. I couldn't get the latest version 2 to work on my 17" isight at all.

MTR site is down right now do to legal issues but if you look in the right places or ask the right questions you can get ahold of him.

If you already have V3 then I don't know what to say.

Anyone know if there is a riplock on the Matshita UJ846? Some companies have been known to lock the rip speed of a drive in the drives bios and an unofficial flash is required to get around it. Never heard of Matshita doing it though.

Let the machine settle in for a week or two and then post again.

Italoop Author

Level 1

15 points

Nov 24, 2005 6:58 AM in response to Rod Hagen

Still not sure what speed you were running the iMac
at teh store in, during these tests, Italo.

Have a look at
http://www.macintouch.com/perfpack/comparison.html
for an example of the quite dramatic differnce in
many tasks between using "automatic" and "highest"
processor speed on the iMac. As you will see, the
mini does, in fact, outperform the iMac when the
iMac is in "automatic" mode, but is well and truly
walloped when the iMac is set to "highest".

Drop me an email for some thoughts on the DVd issues
- my address is in my profile.

Cheers

Rod

Hi Rod

WOW!
Those tests are a revelation. Pretty much what I have been experiencing so far.
All Imacs in the store were personally set to run at Max.
I'll email you.

Thanks a bunch
ItaLOOP

Italoop Author

Level 1

15 points

Nov 24, 2005 7:11 AM in response to Kerr

I may have the answer for mac the ripper. If you are
not using the version 3 beta (I don't think three is
fully out yet) then you are lucky that MTR is
working on the g5 at all. I couldn't get the latest
version 2 to work on my 17" isight at all.

Hi Kerr
I'm using V2.6.6 on the IMac Isight 20".
Works fine here ! (?)

Let the machine settle in for a week or two and then
post again.

mmmh... I am not sure this is the way machines work...
Please read Ron link. That'll show you a lot of good info.

Cheers
I

Rod Hagen

Level 7

32,000 points

Nov 24, 2005 8:26 AM in response to Italoop

mmmh... I am not sure this is the way machines work...

Please read Ron link. That'll show you a lot of good info.

Hi, Italo,

Its worth remembering that the Macintouch test was undertaken with the "first generation" iMac G5, which has a slower processor, and some other significant differences in design, than the present one. We also simply don't know whether it was "straight out of the box" or not either. My own experience with Macs under recent versions of OSX is that you do actually get some speed improvement over the first few days after a "clean" OS installation.

Still, as the Macintouch article indicates, unless you use the iMac with its processor set to highest, in the test the mini does very well by comparison.

Cheers

Rod

Italoop Author

Level 1

15 points

Nov 25, 2005 7:16 AM in response to Rod Hagen

I thought you might be interested in the last evolution of the story....

Well...
I had a lot of curiosity about the HD demo trailers and similar material available on the Apple website.
At the shop we downloaded a pretty hefty amount of them...1.16 GB! This includes King Kong and The Da Vinci Code trailers and all the HD Quicktime Gallery demos available here:
http://www.apple.com/quicktime/guide/hd/
Obviously the shop had all the "crunchiest" Macs to enjoy these beautiful videos, including:
IMac 20"
IMac 17"
Power Mac 2x 2.7 GHz w/ATI 9650 256MB ram
Cinema Display 20", 23" and 30"
...what better place to test VIDEO on a Mac than this?
It turns out...after having downloaded and stored these video files...that many of them "stutter"....the video seems to be "jumpy". This is particularly true in the Da Vinci Code trailer when, at the beginning, the observator "camera" is flying into the cracks and later on, when the DA VINCI CODE text explodes in pieces and they fly to the right side of the monitor....
We tried all available demo resolutions...even the big 2x2.7GHz G5 had a hard time.
The best performace, meaning the stutter artifacts where less prominent, showed up on the IMac 17" G5.
The 20" is pathetic. The big 2x2.7GHz dude lags...seriously.
We tried all the 3 Cinema displays....obviously the 17" of the IMac helped A LOT! Smaller screen and resolution.
The King Kong trailer showed nasty "digital fog" artifacts on mid grey halos on both IMacs...really poor performance.
In both trailers the texture of the ATI X600 is very poor, compared to the 9650!
But still...the 9650 w/256MB...lags and stutters.
One would think that a later generation X600 would be ok.
Nope.
We watched the trailers ALWAYS in their real size in Quicktime...and in full screen mode. No difference at all. Same artifacts.
I was very disappointed as my PCs had no problem at handling any of these demos:
1- P4 3.4 GHz socket 478/2 Gig ram/ GeForce FX5900XT agp
2- P4 3.2 GHz socket 775/1 Gig Ram/ ATI X700 Pci Ex
both showed no artifacts at all.
Surprised? I am!
Thoughts:
-maybe Apple should consider to equip the IMacs w/an ATI X700 at least. DVDs are not nice on the current lower X600.
-it's frustrating one is practically forced to get a Power Mac w/GeForce 6600 (same level graphic card as the Ati X700) or upgrade to a 6800 or even a 7800...just to have the performance you expect from any mid_range pc today.
Consider the price of the system.....it's scary!
Acts:
I gave back my new IMac 20" G5 after 48 hrs with sorrow for the beauty of the object...and great relief for the poor performance it showed.
Placed an order on a Power Mac 2x2.0 GHz G5 w/2.5GB ram and an FX 6800 or 7800 (depending on dealer best offer).
The monitor will not be a Cinema one. It makes no sense.
HP has a 23" model with the same exact LG/Philips panel as the Apple...with DVI-VGA-Video-S_Video and...check this out...COMPONENT! Being a professional designer of surround audio algorithms on hardware DSP products, I often use DVDs to test audio on a professional 5.1 system. Being able to connect the DVD player at 1080i resolution thru Component will be so juicy!!! For 100 Euro less...and 3 years warranty, pivot mode, PiP capability, UI parameters available and input switch!
Please Apple...time to get real with your beautiful products! Really!!!
Honestly I found ridiculous Apple doesn't include wireless/bluetooth on these dual processors machines !!! They are practically anywhere else but them and the IBooks....
Well..anyway....I'm forced to an escalation in price to get expected honest performance from an Apple.
I hope it will deliver or it'll be it w/Macs. Happy w/the Mini for what it does...that's all.

Love the Mighty Mouse though! 😉

Thanks everybody and Rod for the good advices that led me to understand the limitations I encountered.

all the best
Italo

Italoop Author

Level 1

15 points

Nov 25, 2005 7:31 AM in response to Italoop

It turns out...after having downloaded and stored
these video files...that many of them
"stutter"....the video seems to be "jumpy". This is
particularly true in the Da Vinci Code trailer when,
at the beginning, the observator "camera" is flying
into the cracks and later on, when the DA VINCI CODE
text explodes in pieces and they fly to the right
side of the monitor....

Ok ...a correction here.
The sequence in the Da Vinci Code is actually when Monna Lisa face explodes in pieces. There is no text exploding.

Psst... testing right now, out of curiosity
my Toshiba laptop, P4 2.8GHz/1Gig Ram/slow HD/GeForce FXGo5600 w/64MB ram works smooth with the trailer!
Better than all those Macs I tested.
Why is this so? I would expect it to only beat the MINI.

I don't know....

cheers
I

Nov 25, 2005 7:36 AM in response to Italoop

The test on macintouches page measures the older 1.8 ghz iMac (rev A?). Hopefully, if they did it again, the would be some minor performance gains from the iSight model.

Glor

Nov 25, 2005 7:53 AM in response to Italoop

I just downloaded the King Kong trailer and played it on my iMac. I got the 720p version since this is the largest Apple recommends for anything short of a dual processor machine. It played flawlessly in both full screen and normal resolution. I saw no artifacting or stuttering of the video and I was in automatic mode.

Glor

Italoop Author

Level 1

15 points

Nov 25, 2005 7:56 AM in response to Glorfindeal

I just downloaded the King Kong trailer and played it
on my iMac. I got the 720p version since this is the
largest Apple recommends for anything short of a dual
processor machine. It played flawlessly in both full
screen and normal resolution. I saw no artifacting or
stuttering of the video and I was in automatic mode.

Glor

Hi Glor

please read my posting carefully.

King Kong showed "ghosting" artifacts, no stutter.
Da Vinci Code shows stuttering.

cheers
I

This thread has been closed by the system or the community team. You may vote for any posts you find helpful, or search the Community for additional answers.

Mini Mac crunches IMac G5! I'm seriously deluded!