Registry Cleaner Review and Performance Comparison Study

By Jouni Flemming and Veera Peltonen, both working for Macecraft Software, the author of jv16 PowerTools.

Introduction

There are dozens of different Windows registry cleaning related products and nearly all claiming to fix or clean the registry better than the
others. However, there are virtually no real comparisons between these products especially focusing on their actual registry cleaning or error fixing performance.

We have done the following registry cleaner comparison study between the most popular registry cleaners on the market.The selection of products was based on the writers’ experience on which products are the most popular products.

Featured products

Product Version Setup size Installation size Comes with bundled software Price Product type
CleanMyPC Registry Cleaner 4.12 2.8 MB 5.9 MB
No
$29.95 Standalone registry cleaner with some additional features
EasyCleaner 2.0 2.8 MB 3.3 MB
No
$0
(Freeware)
System utility suite including a registry cleaner
jv16 PowerTools 2008 1.8.0.468 6.9 MB 9.0 MB
No
$29.95 System utility suite including a registry cleaner
jv16 PowerTools 2009 1.9.0.528 5.2 MB 6.4 MB
No
$29.95 System utility suite including a registry cleaner
Registry First Aid 7.0.0.1648 3.6 MB 7.9 MB
No
$27.95 Standalone registry cleaner
Registry Mechanic 8.0.0.900 7.1 MB 9.5 MB
Yes
(Google toolbar)
$29.95
per year
System utility suite including a registry cleaner
RegSeeker 1.55 0.4 MB 0.6 MB
No
$24.90
(Free for personal use)
Standalone registry cleaner
Wise Registry Cleaner Pro 3.9.0.143 1.6 MB 2.5 MB
No
$19.95 Standalone registry cleaner

All the products were downloaded from the website of their manufacturer, not from
third party downloading sites. All tests were performed using the trial version of the products
or the full version in the case of freeware products. All the products were installed to the
test setups one by one, started once, then closed and all automatically starting parts of the products were disabled
in order to prevent them from interfering with the other products. After all the products were
installed the test computer was restarted and it was verified that no product had any automatically
starting components started. After this the tests were performed. All timings were done three times and the average time is used.

Basic comparison

These tests were performed on 5th of February 2009.
The tests were performed on a typical Windows Vista Home setup with 1.6 GHz Intel
processor and 2 GB of RAM. The test computer’s registry did not contain any testing
material or created errors, only normal data.

Product Starts automatically Startup time Memory usage Thread usage Registry scan time Found registry errors
CleanMyPC Registry Cleaner
Yes
14 seconds 5 MB 2 2 minutes 41 seconds 292
EasyCleaner
No
3 seconds 2 MB 1 3 minutes 11 seconds 146
jv16 PowerTools 2008
No
9 seconds 19 MB 15 2 minutes 10 seconds 868
jv16 PowerTools 2009
No
4 seconds 15 MB 16 4 minutes 44 seconds 926
Registry First Aid
Yes
4 seconds 6 MB 2 6 minutes 30 seconds 901
Registry Mechanic
Yes
5 seconds 14 MB 3 5 seconds 23
RegSeeker
No
3 seconds 3 MB 1 15 minutes 57 seconds 713
Wise Registry Cleaner Pro
No
3 seconds 11 MB 9 5 minutes 46 seconds 213

Registry cleaning engine accuracy comparison

These tests were performed on 6th of February 2009.
The tests were performed on virtual machine using an installation with minimal third party software installed of
most up-to-date version of Windows 2000. Windows 2000 was selected as the Operating System because it creates
so few registry entries itself to the registry. The computer was a HP Microtower PC with Intel Core 2 Duo processor and 3 GB of RAM
(of which 1 GB was assigned to the virtual Windows 2000 test environment). Most recent
version of VirtualBox was used as the virtualization system. As before, all registry cleaner
products were run with their default settings.

The test environment’s registry was manually modified to contain 20 registry entries of test data, 10
entries to test for false positives (i.e. registry data that might look like invalid but is in fact valid) and 10
entries to test for positives. Set of only 20 entries was selected because some of the tested products did not contain
such basic user interface features as a Search within the results or the use of Search words. Without either of these
two features the analysis using a larger test data would be tedious.

Product False positives Positives
CleanMyPC Registry Cleaner
0
0
EasyCleaner
0
6
jv16 PowerTools 2008
0
10
jv16 PowerTools 2009
0
10
Registry First Aid
0
10
Registry Mechanic
1
8
RegSeeker
0
6
Wise Registry Cleaner Pro
0
5

How to read the results: The higher the Positive count the better, the smaller the False Positive count the better.

Registry cleaning engine accuracy comparison, round 2

These tests were performed on 8th of February 2009.
The tests were performed with the same virtual setup as the previous test (i.e. Windows 2000 with
all the updates and Service Packs with minimal amount of third party software installed).

The test environment’s registry was manually modified to contain 2000 registry entries of test data, 1000
entries to test for false positives (i.e. registry data that might look like invalid but is in fact valid) and 1000
entries to test for positives. Only the products that contain a way to search for or filter the results were
used.

Product False positives Positives
CleanMyPC Registry Cleaner
0
0
EasyCleaner n/a * n/a *
jv16 PowerTools 2008
0
781
jv16 PowerTools 2009
0
1150
Registry First Aid
106
961
Registry Mechanic n/a ** n/a **
RegSeeker n/a *** n/a ***
Wise Registry Cleaner Pro n/a **** n/a ****

How to read the results: The higher the Positive count the better, the smaller the False Positive count the better.

* = EasyCleaner does not seem to contain any feature to filter or search for strings within the results, therefore it wasn’t possible
to count the results accurately. However, browsing through the result list indicated that the product didn’t seem to find any of the
manually inserted test errors (false positives nor real errors).

** = Registry Mechanic does not seem to contain any feature to filter or search for strings within the results, therefore it wasn’t possible
to accurate count the results. However, browsing through the result list indicated that the product found numerous false positives (and correct
errors as well, for that matter).

*** = RegSeeker does not seem to contain any feature to filter or search for strings within the results, therefore it wasn’t possible
to accurate count the results. However, browsing through the result list indicated that the product found numerous correct errors and 9 false positives.

**** = Wise Registry Cleaner Pro does not seem to contain any feature to filter or search for strings within the results, therefore it wasn’t possible
to accurate count the results. However, browsing through the result list indicated that the product found numerous correct errors and 14 false positives.

Stress test

These test were performed on 6th of February 2009.
The tests were performed with the same virtual setup as the previous test (i.e. Windows 2000 with
all the updates and Service Packs with minimal amount of third party software installed).

This time the test environment’s system registry was flooded with 10 000 registry errors and pieces of junk data.
The purpose of the tests was to see whether the products can handle a very large amounts of registry errors and whether
it slows down the scanning or analysing process more on other products than on others.

Product Registry scan time Found registry errors
CleanMyPC Registry Cleaner 59 seconds
130
EasyCleaner
50 seconds *
74
jv16 PowerTools 2008 208 seconds
4788
jv16 PowerTools 2009 191 seconds
4783
Registry First Aid
5+ minutes **
10010
Registry Mechanic 9 seconds
2179
RegSeeker
5+ minutes ***
472
Wise Registry Cleaner Pro 63 seconds
462

How to read the results: The higher the Found registry error count the better, the shorter time it took to analyse the better.

* = After 30 seconds in to the scanning the user interface stopped giving any indication of progress, Task Manager showed the product using 99% of all processor power.
After 20 seconds of this the product said it was done. The lack of progress and the low number of errors found seem to suggest that some kind of self-triggered
abortion of the scan took place.

** = After 5 seconds in to the scanning the product froze completely and there was no indication of any progress.
The product’s user interface did not respond. According to Task Manager the product was using 99% of
all processor power. After 50 seconds the normal operation continued. Again after 2 minutes of scanning the program stopped reporting any progress,
the progress indicator was stuck on 2% of the second step. After five minutes the scan was aborted because it was not clear whether the product had
frozen completely or whether something was in fact happening.

*** = After 45 seconds in to the scanning the product reported 100% progress of the scan yet still the Stop button was active and the
number of found errors kept increasing. The product’s user interface was responding very slowly. According to Task Manager the product was using between 0% and 90% of
processor power, the amount of processor power used seemed to change in intervals of a few seconds.
After five minutes the scan was aborted because it was not clear whether the product had frozen completely or whether something was in fact happening.

Conclusions

Based on the presented data it seems very clear that jv16 PowerTools 2009 provides the very best registry cleaning in both performance and accuracy
of the products tested in this comparison study. However, this study only consists of a few tests, many more tests could be done for example in
the areas of the registry error fixing accuracy, the detection of different registry error types and the fixing accuracy of non-file reference type of
errors.

We are hoping that this study would inspire independent and non-affiliated persons to further compare and analyse
registry cleaner products and publish the results online. For this purpose the writers are willing to offer their expertise available (free of charge) to any
third party interested in doing similar experiments.

You can discuss this study in our discussion forum.

If you found this study is interesting, you can share it with others using these free tools:

All the studies we have done on registry cleaners:

Terminology

  • Setup size means the size of the product’s installation file, such as setup.exe.
  • Installation size means the size of the product’s installation directory with all the included files.
  • Comes with bundled software means whether the product attempts to install any third party software, such as browser plugins or toolbars.
  • Starts automatically means whether the product with default settings
    configures user’s system in a way that the product is always automatically started with the system.
    This is considered as bad practice because each automatically starting program slows down the system’s startup time
    and consumes memory and processing power while running in the background.
  • Startup time means how long does it take for the product to fully start (i.e. to open its main window).
    The faster the product starts the better.
  • Memory usage means the amount of memory the product uses right after being launched. The memory consumption is measured with the Windows
    Task Manager. The less memory the product uses the better.
  • Thread usage means the number of threads the product uses right after being launched.
    A thread is a separate program module that runs on the background and can be used for example to update the product screen. The thread number is measured
    with the Windows Task Manager. The more threads the product uses the better it will perform with modern multi-core processors, therefore the more threads
    the product utilizes the better.
  • Registry scan time means the time it takes for the product to scan the entire registry (using default settings) from the moment Start button
    is pressed to the moment all results are printed to the screen and ready to be accessed by the user. The faster the product scans the registry the better.
  • Found registry errors means the number of registry errors the product reports to find. This test does not study whether the found errors are in fact valid
    or are there false positives included, we will publish a second review with this information included in the near future. Therefore, the more
    registry errors the product finds the better.
  • False positive means a piece of data that is listed by a registry cleaner product as an error but which is not in fact erroneous. The
    concept of false positives is critical in estimating the safety and thus validity of registry cleaner products because the more false positives a
    product lists the more likely the product can cause computer problems instead of solving them.
  • Positive means a piece of data that is listed by a registry cleaner product as an error and which is erroneous based on some
    objective definition. Classical example would be a file reference that references a file that doesn’t exist.